I have been using OpenAI's GPT-4 API for various tasks such as rewriting Titles, H1s, finding content optimization opportunities, producing multi-shot content briefs etc.
I'm curious how are you leveraging the same?
A constructive forum to share and discuss about the latest in SEO news, updates and resources
Mihir, we've been having some trouble getting our AI to understand character counts and limits when it comes to creating metadata. I'm not on the project, only just learned of it recently, but I don't know that we're using ChatGPT. Did you have any difficulty in getting it to follow those instructions, or was a well-written prompt enough to find success?
Using it to create more helpful and digestible tips for users based on our data (e.g. highlight key takeaways from hotel reviews to help users choose and book the hotel that suits them best)
Have been playing with it to rewrite old content from specific perspectives - then copy edit and fact check but can't beat it as a first draft tool
I use it mostly for Named Entity Recognition (NER), feeding a private GPT with my own documents, which can be, for instance, a very long list of keywords based on a few tens of seed queries; or using clients' PDFs about products/lines of business (this especially if case I don't know a heck about the client's niche, as in one case that produces and distributes industrial cables).
The main use csaes I've had revolve around troubleshooting certain code e.g. schema markup, as well as getting suggestions of how to make content more comprehensive. I tend to stay away from using it for things like H1 or title optimisations, since I leverage my keyword research on those.
Hi Mihir! Likewise :)
On one hand, I use https://numerous.ai/ within google sheets for bulk optimization in scenarios like:
Optimizing/rewriting titles, meta descriptions, headings, FAQs, products descriptions that were known to be suboptimal following new criteria.
Initial translation of copy, and different content areas following certain constraints (that will be later edited/validated by humans)
Competition and industry research, to obtain input again in bulk that can be leveraged for content writing and optimization
Content quality, sentiment, intent, helpfulness assessment. For example, I created the GPT: https://chat.openai.com/g/g-WxhtjcFNs-content-helpfulness-and-quality-seo-analyzer
I've shared more about it with scenarios, examples and resources in presentations like:
https://speakerdeck.com/aleyda/embracing-ai-in-seo-how-to-10x-your-seo-leveraging-ai-bots-number-brightonseoÂ
What's fundamental is though is to maximize the quality output by:
Highly descriptive prompts with requirements and constraints: https://www.aleydasolis.com/en/search-engine-optimization/ai-prompts-digital-marketing-seo-generator/
A good validation system within your SEO/optimization/content workflow to avoid errors/issues
Understanding that for content, if everyone is producing it at scale using the same systems, the key is differentiation, and using human expertise and creativity to achieve it. There's only so far you can go with automated content.
Looking forward to read more about the scenarios others have found useful and leveraging too!